The Lecturers’ Experiences and Generative AI in Developing Communicative Speaking Materials in English Education Program

In classrooms around the world, teachers face a constant challenge: how to help students learn to speak a new language not just correctly, but meaningfully. For decades, the most effective approach has been Communicative Language Teaching, a method that prioritizes real interaction over memorizing grammar rules. The goal is to create situations where learners must exchange information, solve problems, or share opinions, much like they would in daily life. This approach relies heavily on the teacher's ability to design activities that feel authentic and relevant to the students' own lives. Today, a powerful new tool has entered the scene: generative artificial intelligence. These are computer systems capable of creating original text, stories, and conversation scripts in seconds. While they offer a way to produce endless teaching materials instantly, they also raise a fundamental question for educators. Can a machine understand the subtle needs of a specific group of students, or does it risk producing lessons that are grammatically perfect but socially hollow?

A team of researchers set out to answer this question by observing how five university lecturers in Indonesia navigate this new landscape. These educators teach English to students who are training to become teachers themselves. The researchers wanted to see how these professionals use artificial intelligence to create speaking materials without losing the human touch that makes language learning effective. They conducted in-depth interviews, asking the lecturers to describe their daily work, their fears, and their strategies when using these digital tools. The study reveals a clear pattern: the lecturers do not let the machines take the wheel. Instead, they treat artificial intelligence as a helpful assistant that generates raw ideas, while the human teacher remains the expert who shapes, edits, and approves the final lesson.

The lecturers in the study described a workflow where the human element is non-negotiable. They explained that while textbooks are useful, they often fail to capture the specific interests, cultural backgrounds, or future career goals of their students. A teacher might need a speaking activity about local tourism in Bone or environmental issues in South Sulawesi, topics that a generic textbook might miss. When the lecturers use artificial intelligence, they typically start by asking it for a list of ideas or a rough draft of a conversation. They might ask for a role-play scenario or a debate topic. However, they almost never use the output exactly as it appears. One lecturer noted that the computer might produce a conversation where everyone is perfectly polite and understands each other instantly, a scenario that rarely happens in real life. Real conversations are messy; people interrupt, hesitate, and sometimes misunderstand one another. The lecturers take the computer's draft and rewrite it to include these natural imperfections, ensuring the material feels authentic.

This process of editing is where the teacher's professional judgment shines. The researchers found that the lecturers act as a filter, checking the computer's work for cultural relevance and practicality. They ask themselves if the vocabulary is too difficult for their specific students or if the situation described would actually make sense in an Indonesian classroom. They also consider the emotional side of learning. If a teacher corrects a student's grammar every time they speak, the student may become too afraid to talk. The lecturers in the study explained that they often let students speak freely first to build confidence, saving corrections for later. A computer, they noted, does not inherently know when to hold back on corrections to keep a student engaged. The human teacher knows when to intervene and when to let the conversation flow, a decision based on years of experience with the specific group of people in front of them.

The study also highlights that the lecturers view these tools as a way to overcome creative fatigue. Teaching the same course semester after semester can be exhausting, and the lecturers admitted that the computer helps them generate fresh ideas quickly. It can suggest a dozen different ways to practice a specific skill in a matter of seconds, giving the teacher a wider menu of options to choose from. However, the lecturers are careful to maintain their authority over the final product. They see the computer as a tool that handles the routine work of brainstorming and drafting, freeing them up to focus on the deeper work of pedagogy. One lecturer described the relationship simply: the computer provides the draft, but the teacher produces the final teaching material. This distinction is crucial. The technology speeds up the process, but it does not replace the need for a human to decide what is best for the learners.

Perhaps the most significant finding is that the lecturers are actively teaching their students to be critical of these tools as well. The researchers observed that the educators do not just use the technology themselves; they encourage their students to question it. In class, they might ask students to compare a conversation generated by a computer with a real-life recording or a textbook example. They discuss how the computer might make mistakes, repeat certain patterns, or miss cultural nuances. This approach turns the use of artificial intelligence into a learning opportunity in itself. Students learn that while the technology is powerful, it is not neutral or infallible. They are taught to verify information and to recognize that the computer's output is just a starting point, not the final truth.

The research concludes that the most successful integration of artificial intelligence in language teaching happens when the technology supports, rather than replaces, human expertise. The lecturers in this study found a healthy balance where the computer handles the heavy lifting of idea generation, and the human teacher ensures the material is meaningful, culturally appropriate, and pedagogically sound. The study suggests that as long as teachers remain the ones making the important educational decisions—setting goals, adapting to student needs, and judging the quality of interaction—the technology can be a powerful ally. It is not a matter of choosing between human teachers and machines, but rather finding a way for them to work together, with the human always guiding the direction. The future of language teaching, according to these educators, relies on keeping the human judgment at the center, using the machine to enhance, not erase, the art of teaching.